Cancer and Metastasis Reviews

, Volume 32, Issue 1, pp 77–82

Mouse models of lung squamous cell carcinomas

Authors

  • Michael S. You
    • Department of Surgery and The Alvin J Siteman Cancer CenterWashington University School of Medicine
  • Lucina C. Rouggly
    • Department of Surgery and The Alvin J Siteman Cancer CenterWashington University School of Medicine
  • Ming You
    • Department of Pharmacology and ToxicologyMedical College of Wisconsin
    • Department of Surgery and The Alvin J Siteman Cancer CenterWashington University School of Medicine
    • Department of Surgery and The Alvin J. Siteman Cancer CenterWashington University
Article

DOI: 10.1007/s10555-012-9406-4

Cite this article as:
You, M.S., Rouggly, L.C., You, M. et al. Cancer Metastasis Rev (2013) 32: 77. doi:10.1007/s10555-012-9406-4

Abstract

Although many mouse models of lung adenocarcinoma exist, only a few mouse lung squamous cell carcinoma models have been developed. Since most clinical chemoprevention trials of lung cancer are performed in subjects with bronchial dysplasia, development of a lung squamous cell carcinoma mouse model sufficient for chemoprevention studies is a high priority. We have shown that lung squamous cell carcinomas can be induced chemically in several strains of mice (1), and that this chemically induced lung squamous cell carcinoma model is applicable to cancer chemoprevention studies. Recently, Ji et al. (2) have shown that simultaneous activation of KrasG12D and inactivation of Lkb1 results in a broader histological range of lung tumors, with approximately 50 % of the lung tumors being squamous cell carcinomas. Here, we review the application of mouse lung squamous cell carcinoma models with different stages of squamous lesions and squamous cell carcinomas to cancer development and chemoprevention studies.

Keywords

Mouse modelsLung squamous cell carcinomaChemical-inducedTransgenicChemoprevention

1 Introduction

Lung cancer is the leading cause of cancer deaths in men and women in the USA [3]. Epidemiological and laboratory animal model studies have demonstrated that smoking is closely linked to increased lung cancer risk [47]. Tobacco exposure has been implicated in 90 % of lung cancers, and smokers have a twentyfold greater risk of developing lung cancer compared with persons who have never smoked [8]. As many as 50 % of all lung cancer cases occur in former smokers and lung cancer is positively correlated with pack years smoked [9, 10]. The percent of smokers in America peaked at almost 50 % in the late 1960s and currently is at 26 % [11]. Although about half of all people who had ever smoked are now former smokers, many people are unable or unwilling to stop smoking. For these reasons, chemoprevention is a potentially important approach to reduce the large number of tobacco caused cancer deaths, especially for former smokers.

Chemoprevention is the use of pharmacologic or natural agents to inhibit the development of cancer. A primary mode of chemoprevention includes reversing the progression of premalignant cells by inhibition of cell cycle progression and induction of apoptosis. Numerous studies have found that chemoprevention methods can prevent or improve the outcome of a wide variety of cancers [12]. This approach is especially useful in targeting people that are at high risk for developing cancer, such as patients who have a genetic predisposition to cancer, or patients who are at high risk of developing secondary primary tumors after surgical resection of a tumor [12]. Pathological stages including hyperplasia and dysplasia are targeted by chemoprevention.

There are four major types of lung cancer: adenocarcinoma (32 %); squamous cell carcinoma (SCC; 29 %); small cell carcinoma (18 %); large cell carcinoma (9 %); and uncommon and poorly differentiated types comprising the remaining 12 % [13]. Because of similarities in histopathology and tumor progression stages between mouse and human lung adenocarcinomas, mouse lung adenoma, or adenocarcinoma models have been used extensively to evaluate the efficacy of putative lung cancer chemopreventive agents [5, 14]. However, only a few mouse lung squamous cell carcinoma models have been developed. Most lung squamous cell carcinomas arise centrally from either the main, lobar, or segmental bronchi and ulcerate through the mucosa into the surrounding lung parenchyma [13]. Histological and cytological studies have revealed a series of changes that occur over many years and represent a morphological progression to bronchogenic carcinoma [1518]. Early changes include basal cell hyperplasia followed by a squamous metaplasia, dysplasia, carcinoma in situ, and invasive SCC [1518]. There is strong evidence that tobacco smoke plays a major role in the pathogenesis of lung cancer including lung SCC [19]. Neoplasms originating from cells lining bronchi and bronchioles are among the most common pulmonary cancers in humans, and mortality from cancers caused by tobacco products will continue to be significant in the foreseeable future [20, 21].

It has been difficult to establish a mouse lung SCC model (2225; and reviewed by 5, 26). Previous studies, now almost 30 years old, showed a high incidence of mouse SCCs induced by direct tracheal instillation of benzo(a)pyrene (77 %) or 3-methylcholanthrene (85 %) [2224]. These studies were technically difficult, and the results were rarely reproduced by others. Subsequently, a report showed that preneoplastic and neoplastic lesions were induced by skin painting of N-nitroso-trischloroethylurea (NTCU) and of N-nitroso-methyl-bischloroethylurea [25]. However, neither of these two mouse lung SCC models has been fully characterized or routinely used for mechanistic, genetic, or preclinical cancer therapeutic or chemoprevention studies.

One of the important factors for some of the failures in recent clinical trials in subjects with bronchial dysplasias can be attributed to the lack of appropriate animal models for human squamous cell carcinomas. As discussed above, the A/J mouse lung adenoma model is similar histologically and at the molecular level to a subtype of human adenocarcinoma and has been widely used in studies of the chemoprevention of lung adenomas or lung adenocarcinoma [5, 26]. However, there is no validated mouse lung squamous cell carcinoma model for use in preclinical cancer chemoprevention studies. To this end, we recently developed and validated a mouse model of lung squamous cell carcinomas [1].

2 A chemically induced mouse model

Tumor induction and histopathology

Mouse models for lung adenoma and adenocarcinoma are well-established, but a mouse model for lung SCC has been lacking. We are one of the few groups that have extensive experience using a mouse lung SCC model in cancer chemopreventive studies [1, 27, 28]. We established a human lung SCC model in mice by treating mice with NTCU by skin painting. Specifically, 6- to 8-week-old mice were treated topically with NTCU twice a week for a period of 7 to 9 months. After termination, the lung was removed and fixed in 10 % buffered formalin for 48 h and then transferred to 70 % ethanol until embedded in paraffin. In our earlier study [1], we treated eight strains of mice including 129/svJ, AKR/J, BALB/cJ, C57BL/6J, FVB/J, SWR/J, A/J, and NIH Swiss with NTCU. This treatment regimen is sufficient to induce lung tumors in certain strains of mice but not in others. Lung SCC was seen in five of the eight strains tested, whereas the lung SCC in situ was seen in seven of eight strains with NTCU treatment. Premalignant lesions, including lung hyperplasia and metaplasia, were seen in all of mouse strains tested. NIH Swiss, SWR/J, and A/J mice had a greater neoplastic burden compared with FVB/J or BALB/cJ mice. The lower susceptibility of AKR/J, 129/svJ, and C57BL/6J mice might appear partially because of fewer mice treated. However, these differences are readily apparent when considering the high multiplicity of tumors in the various susceptible strains. Lung SCCs were observed in 75–100 % of NIH Swiss, SWR/J, and A/J mice. An SCC tumor incidence of 40–45 % was observed in BALB/cJ and FVB/J mice, respectively. AKR/J, 129/svJ, and C57BL/6J mice are resistant to lung SCC induction by NTCU skin painting. An adenoma, with adenocarcinoma arising in its center, was observed in two of eight BALB/cJ mice and two of nine FVB/J mice (data not shown). Based on these strain-specific differences in susceptibility to mouse lung SCC, a whole-genome linkage disequilibrium analysis was conducted in seven strains of mice, divided into three phenotype categories of susceptibility, using Fisher’s exact test applied to 6,128 markers in publically available databases. Three markers were found to be significantly associated with susceptibility to SCC, P < 0.05. They were D1Mit169, D3Mit178, and D18Mit91. Interestingly, none of these sites overlap with the major susceptibility loci associated with lung adenoma/adenocarcinoma development in mice.

For histopathology analysis, more than 100 serial tissue sections (5-μm each) were made from formalin-fixed lung, and one in every 20 sections (approximately 100 μm apart) was stained with H&E and examined histologically under a light microscope. The lesions, including invasive SCC, SCC in situ, and bronchial hyperplasia/metaplasia, were scored from the H&E-stained sections of each lung. When hyperplasia occurs (Fig 1b), a single layer of bronchiolar epithelial cells (Fig 1a) are transformed into multiple layers. The cells maintain their normal appearance. Mitosis is rare. In bronchiolar metaplasia (Fig 1c), the normal columnar epithelium is replaced by flattened squamous epithelium with increased keratin production (increased red staining in H&E). The premalignant lesions described in this paper are a combination of bronchiolar hyperplasia and bronchiolar metaplasia. In carcinoma in situ (Fig 1d), atypical cells (irregular shape, increased nucleus/cytoplasm ratio) with mitotic figures and loss of orderly differentiation replaced the entire thickened epithelium. The bronchiole basement membrane is intact. There is no tumor cell in the surrounding stroma. In invasive carcinoma (Fig 1e), general features of SCC such as keratin pearl, multiple nuclei, and increasing mitotic index can be seen. The normal architecture of the lung is disrupted. Cords and nests of tumors can be seen in the subepithelial stroma. Immunohistochemical staining for squamous marker, cytokeratin (CK) 5/6, found that CK5/6 is expressed in the cytoplasm of both human and mouse lung SCC cells (Fig 1f). Squamous origin of squamous lesions was confirmed by Hudish TM et al. [29] with IHC staining for cytokeratin 5/6, p63, thyroid transcription factor-1 (TTF-1), and Napsin-A.
https://static-content.springer.com/image/art%3A10.1007%2Fs10555-012-9406-4/MediaObjects/10555_2012_9406_Fig1_HTML.gif
Fig. 1

Mouse lung squamous lesions. a Normal bronchus; b, hyperplasia; c, metaplasia; d, carcinoma in situ; e, squamous cell carcinoma in mouse lung; and f, mouse pulmonary carcinoma with positive staining for CK5/6 (f is adapted from ref. [1]). All photos were taken at ×40 originally

Applications to cancer chemoprevention studies

This mouse lung SCC model is highly important for preclinical studies of lung cancer chemopreventive agents because most human trials have been conducted against precancerous lesions for lung SCC. Previous studies have shown significant chemopreventive efficacy of antitumor B (ATB), a Chinese herbal mixture of six plants, against human esophageal and lung cancers. We have tested the chemopreventive efficacy of ATB in our mouse lung SCC model. Figure 2a showed the general experimental design for testing chemopreventive agents using the chemically induced mouse lung SCC model. We found that ATB decreased lung SCC load significantly (3.1-fold; P < 0.05) and increased lung hyperplastic lesions by 2.4-fold (P < 0.05; Fig. 2b). This observation suggests that ATB can block hyperplasia from progression to SCC. These results indicate that ATB is a potent chemopreventive agent against the development of mouse lung SCCs [27]. In a recent study, we demonstrated that delayed administration of pioglitazone (a ligand of nuclear receptor PPARγ) caused a 35 %, (P < 0.05) decrease in lung SCC [28]. More recently, we tested green tea for its chemopreventive efficacy in A/J mice (Wang et al. unpublished results). Mice were given 0.6 % green tea as their source of water starting 2 weeks after the initial NTCU administration and continuing for the duration of the experiment. All mice were terminated 32 weeks after the initial NTCU administration. We found that 90 % (nine out of ten) of the wild-type mice developed lung SCC in the control NTCU-treated group and only 56 % (five out of nine) in the NTCU-green tea-treated group. To assess lung SCC development, serial tissue sections were made and stained with H&E. Bronchial lesions, including hyperplasia, metaplasia, SCC in situ, and invasive SCC, as well as normal bronchus were scored and converted into percentages. Mice treated with NTCU have an average of 37.6, 18.7, 9.4%, 8.6, and 23.9 % of normal bronchia, hyperplastic bronchia, metaplastic bronchia, SCC in situ, and SCC per lung, respectively. Mice treated with NTCU-green tea have an average of 41.7, 34.4, 6.8, 4.2, and 5.9 % of normal bronchia, hyperplastic bronchia, metaplastic bronchia, SCC in situ, and SCC per lung, respectively. The differences in the percentage of hyperplasias, SCCs in situ, and SCCs are statistically significant between the two groups. Treatment of green tea decreased SCC by 4.1-fold, decreased SCC in situ by 2.1-fold, and thus increased hyperplasia by 1.8-fold (Wang et al. unpublished data). These results indicate that green tea has chemopreventive effects on NTCU-induced lung SCC in wild type A/J mice. Green tea exerts its effect by preventing the progression from the hyperplastic stage to the malignant stage. Green tea may also prevent hyperplastic change from normal bronchia. Studies on the characterization of NTCU-induced lung SCC and utilization of this model in lung cancer chemoprevention are summarized in Table 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs10555-012-9406-4/MediaObjects/10555_2012_9406_Fig2_HTML.gif
Fig. 2

Chemopreventive study using NTCU-induced lung SCC model in mice. a General experimental design. Mice at age ~8 week are treated topically with 0.04 mol/L NTCU (arrow) twice a week for 32 consecutive weeks. The time of first dose of NTCU is counted at time zero. Two weeks after the start of NTCU treatment, mice are fed either AIN-76A Purified Diet #100 000 or the same diet potential chemopreventive agent (arrowhead). Thirty-four weeks after the initial treatment of NTCU, mice are terminated by CO2 asphyxiation. Horizontal lines represent the time (by weeks). Solid line indicates the mice treated with NTCU-only (as diet control, DC). Shaded line indicates the mice treated with NTCU and testing agent(s). b Example of lung cancer chemopreventive study using the SCC model. Efficacy of antitumor B (ATB). Due to the difficulty to establish a tumor count on lung SCC with conventional approach as mouse SCC does not form visible solid nodules on the surface of the lung, serial tissue sections were made from each formalin-fixed lung and one in every 20 sections was stained with hematoxylin and eosin (H&E). To assess specific effects of these agents on each histopathologic stage, all of the bronchial in each given slides were counted and grouped into five categories based on normal, hyperplasia, metaplasia, SCC in situ (dysplasia was included in this category), and invasive SCC. The number in each category was then converted into percentage. ATB, anti-tumor B; DC, diet control; *P < 0.05. (This figure is modified from the reference [27])

Table 1

Summary of the studies on characterization of NTCU-induced lung SCC and utilization of this model in lung cancer chemoprevention

NTCU

Mice

NTCU-induced lung SCC

Reported chemopreventive (CP) study

Others findings

References

Dose

Topical treatment

Strain(s)

Gender

Squamous markers

Alveolar type II cell markers

NE cell markers

CP Agent

dose

Agent Started weeks before or after first dose of NTCU

Effect

0.04 mol/l (molar) in acetone

Twice a week for 35 to 40 weeks

Swiss mice (Cr:NIH(S))

Female

Positive

Negative

Negative

Not tested

    

[32]

4 mmol/L in acetone

Twice a week for 8 months

129/svJ, A/J, AKR/J, BALB/cJ, C57BL/6J, FVB/J, NIH Swiss, and SWR/J

Female

Positive

Not tested

Not tested

Not tested

   

(1) NTCU induces lung SCCs in SWR/J, NIH Swiss, A/J, BALB/cJ, and FVB/J; but not in AKR/J, 129/svJ, and C57BL/6J. (2)D1Mit169, D3Mit178, andD18Mit91were found significantly associated with susceptibility to SCC.

[1]

4 mmol/L in acetone

Twice a week for 240 days

A/J

 

Not tested

Not tested

Not tested

Pomegranate fruit extract

0.2 % (w/v)

1 week before

Positive

 

[33]

4 mmol/L in acetone

Twice a week for 8 months

NIH Swiss mice

 

Not tested

Not tested

Not tested

Not tested

   

18F-FDG-PET is feasible to monitor NTCU-induced lung SCC growth in mice

[34]

4 mmol/L in acetone

Twice a week for 8 months

A/J

Female

not tested

not tested

not tested

Antitumor B

250 g/kg diet

2 weeks after

Positive

 

[27]

3 mmol/L in acetone

Twice a week for 8 months

NIH Swiss mice

Female

not tested

not tested

not tested

Pioglitazone

15 mg/kg BW

8 weeks after

Positive

 

[28]

4, 8, or 40 mmol/L

Twice a week for 32 weeks

FVB

Male and female

Positive

Negative

Negative

Negative

   

40 mmol/L NTCU produced the entire spectrum of dysplasia and SCC, but was associated with poor survival. 4 and 8 mmol/L NTCU were bettertolerated and produced only significant levels of flat atypia.

[29]

Abbreviations: 18F-FDG, 2-[18F]fluoro-2-deoxy-d-glucose; BW, body weight (gram); NE, neuroendocrine; PET, positron emission tomography; SCC, squamous cell carcinoma

3 A transgenic mouse model

The human LKB gene (STK11), a tumor suppressor gene, encodes a serine/threonine protein kinase that is defective in patients with Peutz-Jeghers syndrome (PJS) [30]. A review of the literature reveals a number of somatic LKB1 mutations in lung and colorectal cancers [30]. Recent studies by Ji H. et al. [2] established LKB1 as a critical barrier to pulmonary tumorigenesis by controlling initiation, differentiation, and metastasis [2]. Their studies have shown that Lkb1-deficient tumors demonstrated shorter latency, an expanded histological spectrum including adenocarcinoma, squamous cell carcinoma, and large cell carcinoma, and more frequent metastasis compared to tumors lacking p53 or Ink4a/Arf tumor suppressor genes [2]. LKB1 inactivation is frequent in all human non-small cell lung cancer (NSCLC) subtypes, suggesting similar tumor suppressor roles in humans [2]. All lung tumors from K-ras-mutated mice are lung adenocarcinomas with or without inactivation of p16Ink4a, Ink4a/Arf, or p53. In contrast, 15 of 27 adeno-Cre-treated K-ras Lkb1L/L or Kras Lkb1L/− mice were either SCC or adenosquamous carcinomas, and 2 of 27 lungs harbored large cell carcinomas (LCC).

Squamous tumors from K-rasLkb1L/L or K-rasLkb1L/− mice did not express pro-surfactant protein C (SP-C), a marker of type II pneumocytes and adenocarcinomas, but expressed pan-keratin and p63, markers of SCC [2]. SP-C expression was high and expression of pan-keratin and p63 was low or absent in adenocarcinomas [2]. Therefore, Lkb1 inactivation facilitated tumors of all three human histological subtypes, particularly lung SCC. LKB1 alteration was also seen in human lung SCCs (8 of 42, 19 %) with the predominant lesion (6 of 42) also being single-copy mutation or deletion [2]. The frequent occurrence of a single-copy mutation or deletion of LKB1 in human tumors is consistent with the increased rate of tumor formation observed in KrasLkb1+/− or KrasLkb1L/+ mice compared to K-ras mice. Inactivating mutations of LKB1 are found in all histologic subtypes of human NSCLC [2]. Unfortunately, there is no genetically engineered mouse model that develops predominantly squamous cell carcinoma in lung as reviewed by Farago AF et al. [31].

4 Summary

SCC of the lung is strongly linked with cigarette smoking. Although there were reports of inducing SCCs in the lung by the intratracheal installation of polycyclic aromatic hydrocarbons almost 35 years ago, this procedure has not been examined further or reproduced. In the 1980s, Lijinsky et al. [25] described induction of lung SCC (LSCC) in NIH Swiss mice by painting mouse skin with NTCU for an extended period of time. The effects of NTCU painting in eight strains of mice demonstrated all different pathological stages of squamous cell carcinoma: bronchial hyperplasia, bronchial metaplasia, SCC in situ, and invasive SCC. Thus, we have not only successfully reproduced Lijinsky’s LSCC model in NIH Swiss mice but also extended it to other strains of mice.

It has been recently reported that Kras-Lkb1-deficient mice have an expanded spectrum of lung cancer subtype histological spectrum, including squamous cell carcinoma, large cell carcinoma, and adenocarcinoma [2]. The implications of these findings should extend to research of cancer chemoprevention of lung squamous cell carcinoma. We suggest that development of new genetically engineered mouse lung SCC models will be important to catalyze future progress in lung cancer chemoprevention.

Copyright information

© Springer Science+Business Media New York 2012